Exploring the

Impact of

COVID-19 on

Population

Demographic

GROUP-25

Vikram Kotlo 11608418

vikramkotlo@my.unt.edu

Introduction

Domain: Health and Demographics Analysis

This domain involves analyzing health-related datasets and

demographical data to understand patterns, trends, and factors

influencing public health outcomes.

The objective is to gain insights into various aspects of health,

including life expectancy, mortality rates, disease prevalence,

and the impact of epidemics such as COVID-19.

By looking in to these datasets, we aim to identify correlations,

disparities, and potential interventions to improve global health

outcomes.

Below

Workflow

Explaination

Step 1: Data Collection (Data Gathering):

Obtain relevant datasets from sources like Kaggle, which provide comprehensive health and epidemiological

data.

Step 2: Initial Visualization (Exploratory Data Analysis):

Use tools like D3.js to create initial visualizations, providing an overview of data structure and potential

insights.

Step 3: Data Cleaning (Data Preprocessing):

Employ Python libraries like Pandas and NumPy to clean the datasets, handling missing values, outliers, and

inconsistencies.

Step 4: Refined Visualization (Advanced Visualizations):

Utilize Python libraries such as Matplotlib and Seaborn to create refined visualizations, focusing on specific

variables of interest and uncovering deeper insights.

Step 5: Dashboard Creation (Interactive Reporting):

Utilize platforms like Microsoft Power BI to design interactive dashboards and reports, integrating the refined

visualizations to provide a comprehensive view of the analyzed data.

Step 6: Report Generation (Insights and Conclusions):

Create detailed reports summarizing the analysis findings, insights, and conclusions drawn from the

visualizations and data analysis process, facilitating informed decision-making in public health policy and

interventions.

Data Abstraction-Dataset (Type and

Attributes)

World Data:

Type: CSV

Attributes:

Country (object): Name of the country.

Year (int64): Year of observation.

Status (object): Health status classification.

Life expectancy (float64): Average life expectancy in years.

Adult Mortality (int64): Mortality rate of adults per 1000 population.

Infant deaths (int64): Number of infant deaths per 1000 live births.

Alcohol (float64): Alcohol consumption in liters per capita.

Percentage expenditure (float64): Health expenditure as a percentage of GDP.

Hepatitis B (int64): Percentage of population vaccinated for Hepatitis B.

Measles (int64): Number of reported measles cases per 1000 population.

BMI (float64): Average Body Mass Index.

Population (float64): Population of the country.

COVID-19 Data:

Type: .CSV

Attributes:

Country/Region (object): Name of the country or region.

Confirmed (int64): Total confirmed COVID-19 cases.

Deaths (int64): Total deaths due to COVID-19.

Recovered (int64): Total recoveries from COVID-19.

Active (int64): Active cases of COVID-19.

New cases (int64): New confirmed cases reported.

New deaths (int64): New deaths reported.

New recovered (int64): New recoveries reported.

Deaths / 100 Cases (float64): Percentage of deaths among confirmed cases.

Recovered / 100 Cases (float64): Percentage of recoveries among confirmed

cases.

Data Transformation

Data Transformation Techniques:

Handling missing values, scaling numerical features, encoding categorical features, and

normalizing data.

Handling Missing Values:

Use techniques like isnull().sum() to address missing values in both datasets.

Scaling Numerical Features:

Scale numerical features to a similar range to prevent biases in models.

Encoding Categorical Features:

Encode categorical features into numerical values using techniques like label encoding.

Normalizing Data:

Normalize numerical features to have a mean of 0 and a standard deviation of 1 to improve

model convergence and performance.

Task

Abstraction

Task: Analyzing Health and Demogrphical Data

Target:

Gain insights into global health trends, identify factors influencing health

outcomes, and assess the impact of epidemics like COVID-19.

Actions:

Investigate the relationship between various factors (e.g., GDP, healthcare

expenditure) and life expectancy to understand health outcomes.

Explore adult mortality and infant mortality rates to identify regions with

high mortality rates and potential interventions.

Analyze data on diseases such as measles and hepatitis B to assess disease

prevalence and vaccination coverage.

Evaluate the spread and impact of COVID-19, including confirmed cases,

deaths, and recoveries, to inform public health responses.

Investigate the influence of socioeconomic factors like GDP, education, and

income composition on health outcomes.

Task Abstraction Workflow Diagram

Analyze Life Expectancy Trends

Explore the relationship between life expectancy and factors like GDP, healthcare expenditure, and vaccination

coverage.

Assess Mortality Rates

Investigate adult mortality and infant mortality rates to identify regions with high mortality rates and potential

interventions.

Study Disease Incidence

Analyze data on diseases such as measles and hepatitis B to assess disease prevalence and vaccination coverage.

Investigate COVID-19 Impact

Evaluate the spread and impact of COVID-19, including confirmed cases, deaths, and recoveries, to inform

public health responses.

Examine Socioeconomic Factors

Investigate the influence of socioeconomic factors like GDP, education, and income composition on health

outcomes.

Implementation

using Tools:

Tools Used for Visualization:

D3.js:

Utilized for initial visualization of uncleaned dataset, providing interactive and dynamic visualizations

on webpages(vizhub).

Python (Pandas, NumPy, Matplotlib, Seaborn):

Pandas and NumPy for data manipulation and preprocessing.

Matplotlib and Seaborn for creating static visualizations such as line plots, bar charts, and histograms.

Microsoft Power BI:

Used for creating interactive dashboards and reports, integrating refined visualizations to provide a

comprehensive view of the analyzed data.

Explanation:

D3.js enabled the creation of dynamic and interactive visualizations for exploring the uncleaned dataset,

providing an initial overview of the data's structure.

Python, with libraries like Pandas and NumPy, was primarily used for data cleaning, preprocessing, and

generating static visualizations.

Microsoft Power BI was employed for creating interactive dashboards and reports, allowing for in-

depth exploration and analysis of the refined visualizations.

Results for

Analysis

Initial Visualization Using D3.js Pie Chart

of COVID-19 Cases by WHO Region(pre-

cleaning)

Explanation:

The pie chart provides a clear overview of how COVID-19 cases are

distributed among different WHO regions.

We can see at a glance which regions have the highest number of

confirmed cases, enabling policymakers and health organizations to

allocate resources effectively.

Regions with larger segments may require more attention in terms of

healthcare infrastructure and preventive measures.

Story:

Imagine global health officials convening to discuss strategies for combating

the spread of COVID-19. As they gather around a screen displaying the pie

chart, they immediately notice the significant portion representing the

European region, indicating a high number of confirmed cases.

This prompts discussions on implementing stricter containment measures

and increasing healthcare capacity in Europe.

Similarly, the smaller segments representing regions with fewer cases spark

conversations about sharing resources and best practices to prevent further

outbreaks.

Stacked Bar Chart of COVID-19 Cases by

Country-Using D3.js

Explanation:

The stacked bar chart allows for a detailed examination of COVID-19 cases within

individual countries. By breaking down each bar into segments for confirmed,

deaths, and recovered cases, viewers can see not only the total number of cases but

also the outcomes of those cases. This provides insights into how countries are

managing the pandemic, including their healthcare systems' capacity to treat

patients and mitigate fatalities.

Story:

As policymakers analyze the stacked bar chart, they focus on the disparities in

outcomes among the top 10 countries with the highest confirmed cases. They notice

that while some countries have a high number of confirmed cases, they've also

managed to achieve a substantial number of recoveries, indicating effective

healthcare interventions. Conversely, countries with a high proportion of deaths

prompt discussions on implementing measures to reduce mortality rates, such as

improving access to critical care facilities and vaccine distribution.

After Cleaning Datasets Using Python And

Visualizing Using Matplotlib: Exploring Global

Health Data

Explanation:

Data Cleaning: Removes extra spaces from column names for

consistency.

Summary Statistics: Computes key statistics for numeric columns.

Visualization: Displays the distribution of life expectancy using a

histogram.

Story:

In a health research project, data cleaning ensures consistency in

column names, aiding smooth analysis.

Summary statistics reveal crucial insights into life expectancy trends

worldwide.

Visualizing life expectancy distributions highlights variations, guiding

targeted interventions for global health improvement.

Understanding Global Population Dynamics: Distribution

Analysis

Explanation:

Data Preparation: Ensures consistency in column names for clarity.

Statistical Insight: Reveals key population statistics for analysis.

Visual Representation: Depicts population distribution via histogram

and KDE.

Story:

In a demographic study, data is prepared by standardizing column

names to streamline analysis. Statistical examination unveils crucial

population trends, essential for informed decision-making. Visualizing

population distributions illuminates demographic disparities, guiding

policy interventions for sustainable development.

Exploring Global Economic Patterns: Distribution of GDP

Explanation:

Data Preparation: Cleans column names, ensuring uniformity.

Statistical Analysis: Computes summary statistics for GDP.

Visualization: Displays GDP distribution via a histogram with

KDE.

Story:

In an economic analysis endeavor, data is prepped by

standardizing column names for clarity. Statistical analysis

uncovers insights into global GDP trends, vital for economic

policy formulation. Visualizing GDP distributions highlights

disparities, informing strategies for balanced economic

development.

Analyzing Numeric Relationships: Correlation Heatmap

Explanation:

Data Filtering: Retains only numeric columns for correlation analysis.

Exploratory Analysis: Examines correlations between numeric

variables.

Visual Insight: Presents correlations via heatmap with annotations.

Story:

In a data exploration endeavor, non-numeric columns are filtered out

to focus on quantitative relationships.

Through exploratory analysis, correlations among numeric variables

are investigated, offering insights into interdependencies.

Visualizing correlations via a heatmap aids in identifying patterns and

guiding further analysis for informed decision-making.

Analyzing COVID-19 Trends: Summary Statistics and

Time Series Plot

Explanation:

Statistical Overview: Generates summary statistics for COVID-19 data.

Temporal Analysis: Plots trends of confirmed cases, deaths, and

recoveries over time.

Story:

In assessing the COVID-19 pandemic, summary statistics offer insights

into the overall magnitude and variability of cases.

Meanwhile, the time series plot depicts the progression of confirmed

cases, deaths, and recoveries over time, enabling the observation of trends

and fluctuations.

Through these analyses, policymakers and health authorities can better

understand the impact of the pandemic and formulate effective response

strategies.

Visualizing COVID-19 Metrics on a World Map

Explanation:

Data Loading: Imports the world map shapefile and COVID-19 data.

Data Merging: Merges COVID-19 data with world map data based on country names.

Map Visualization: Plots COVID-19 metrics (confirmed cases, deaths, recoveries) on the world map.

Story:

In mapping the global impact of COVID-19, the world map shapefile is loaded alongside COVID-19

data.

By merging these datasets, a comprehensive view of the pandemic's spread is achieved, allowing for

spatial analysis.

The resulting visualizations depict the distribution of confirmed cases, deaths, and recoveries across

countries, aiding in understanding regional variations and informing targeted response efforts.

Exploring the Relationship

Between Life Expectancy and

COVID-19 Cases

Explanation:

Data Merging: Merges datasets based on the 'Country/Region' column.

Scatter Plot: Visualizes the relationship between life expectancy and confirmed COVID-19

cases.

Story:

By merging datasets on country/region identifiers, a comprehensive dataset is created for

analysis.

Through a scatter plot, the correlation between life expectancy and confirmed COVID-19

cases is explored, offering insights into potential health disparities and vulnerabilities.

This analysis contributes to understanding how demographic factors may influence the spread

and impact of the pandemic.

Examining Life Expectancy and

COVID-19 Cases Across

Development Status

Explanation:

Data Visualization: Displays a scatter plot of life expectancy against confirmed

COVID-19 cases.

Color Representation: Categorizes data points by development status,

enhancing visual interpretation.

Story:

In this scatter plot analysis, life expectancy is juxtaposed against confirmed

COVID-19 cases, with data points color-coded according to development

status.

The visualization facilitates the identification of potential correlations between

socioeconomic factors and pandemic outcomes.

By examining disparities across development statuses, policymakers can tailor

interventions to address health inequities and mitigate the impact of the

pandemic on vulnerable populations.

Exploring GDP and COVID-19

Cases Across Development Status

Explanation:

Visualization: Presents a scatter plot of GDP against confirmed COVID-19

cases.

Color Coding: Categorizes data points by development status for enhanced

insight.

Story:

This scatter plot delves into the relationship between GDP and confirmed

COVID-19 cases, with data points color-coded based on development status.

By examining how economic factors intersect with pandemic outcomes,

stakeholders can better understand the socioeconomic dimensions of the crisis.

This analysis aids in identifying disparities in vulnerability and resilience

across different economic strata, guiding targeted interventions for pandemic

response and recovery.

Comparing Average GDP

Across WHO Regions

Explanation:

Visualization: Presents a bar plot of average GDP by WHO region.

Data Aggregation: Computes the mean GDP for each WHO region for

comparison.

Story:

This bar plot showcases the average GDP across different WHO regions,

providing insights into regional economic disparities.

By aggregating GDP data, the plot highlights variations in economic

development among WHO regions.

Stakeholders can use this analysis to prioritize resource allocation and

development initiatives, aiming to address economic inequalities and

promote sustainable growth globally.

Comparing Average Life

Expectancy Across WHO Regions

Explanation:

Visualization: Displays a bar plot of average life expectancy by WHO region.

Statistical Summary: Calculates the mean life expectancy for each WHO region for

comparison.

Story:

This bar plot illustrates the average life expectancy across various WHO regions, shedding

light on regional health disparities.

By analyzing average life expectancy data, stakeholders can identify regions with lower life

expectancies and prioritize health interventions accordingly.

This analysis aids in understanding global health outcomes and guiding efforts to improve

population health and well-being across different regions.

Microsoft Power BI

Work Management

Work Completed:

Data Acquisition and Cleaning:

Obtained COVID-19, global population, and economic datasets. Cleaned data to ensure consistency and

accuracy.

Initial Visualization Using D3.js:

Created a pie chart of COVID-19 cases by WHO region using D3.js.

Stacked Bar Chart of COVID-19 Cases:

Developed a stacked bar chart showing COVID-19 cases by country using D3.js.

Cleaning Datasets Using Python and Visualization with Matplotlib:

Cleaned datasets using Python. Visualized global health data with Matplotlib.

Exploring Global Population Dynamics:

Analyzed population distributions and trends.

Exploring Global Economic Patterns:

Examined GDP distributions and correlations with COVID-19 cases.

Understanding Numeric Relationships:

Investigated correlations between variables using a heatmap.

Work Management(Cont.)

Analyzing COVID-19 Trends:

Generated summary statistics and time series plots for COVID-19 data.

Visualizing COVID-19 Metrics on a World Map:

Plotted COVID-19 metrics on a world map.

Exploring Relationships Between Life Expectancy and COVID-19:

Investigated the relationship between life expectancy and COVID-19 cases.

Links

Dashboard Link :

https://app.powerbi.com/groups/32943138-8025-48cd-9a85-

9df66bd1864a/dashboards/fab773de-71a6-474f-9117-19982ed0492f?ctid=70de1992-

07c6-480f-a318-a1afcba03983&pbi_source=linkShare

Vizhub code:

https://vizhub.com/Vikramkotlo09/3e4a96b9055a47c6a872c51257676511?edit

=files&file=index.html

Viz hub Visualization:

https://vizhub.com/Vikramkotlo09/3e4a96b9055a47c6a872c51257676511?mod

e=embed

References:

1. World Health Organization. (2022). COVID-19 Dashboard. [Online]. Available:

https://www.who.int/emergencies/disease-outbreak-news/item/2020-DON233

[Accessed: April 16, 2024].

2. In-Class activities gave good knowledge and reference to complete this project.

3. D3.js Documentation: https://d3js.org/

4. Matplotlib Documentation: https://matplotlib.org/stable/contents.html

5. Pandas Documentation: https://pandas.pydata.org/docs/